On the face of it, running a successful recruitment round is fairly straightforward.
It’s a bit like making a salad. Everything needs to look fresh and appetising, and you don’t want too much of one thing in case people don’t like it.
I mean, it’s not rocket science.
The provider level data from UCAS nicely illustrates the other, less straightforward end of the equation. We know surprisingly little about what applicants actually want to do, and where they want to do it.
Sure, there’s near-certainties – medicine at UCL is unlikely to want for well-qualified applicants any time soon – but some things are rather less expected. Computing and IT focused courses, which have been growing in popularity for years, appear to have hit a wall. Is it the onset of generative AI “vibe coding” hitting employment prospects? Is it a change in the public perception of technology companies?
We pretty much know it is affordability (and the slow atrophy of the student maintenance system) that prompts applicants from less advantaged socio-economic backgrounds to choose to study locally. But we don’t know why selective providers that have historically recruited nationally have decided en masse to move into this very specialised market, or what changes they have made to their standard teaching (and indeed offer-making) approach to make this work.
It’s questions like these that make the insights available from this year’s UCAS End of Cycle data so fascinating, and the choice of data that is released so frustrating.
The Russell Group ate my students!
There’s been a lot of talk (and a lot of quite informed data driven evidence) to suggest that traditionally selective providers have been accepting students with uncharacteristically low grades in greater numbers than in previous years.
A couple of unexpected new additional data tables shed a little more light. This last (2025) cycle saw selective (high tariff) providers recruit more students with 15 A level points or below than in any previous year – while medium tariff providers are doing less well in students with between 9 and 11 points than any year outside the pandemic, and low tariff providers had their worst year on record for between 10 and 12 points, and their worst year since the pandemic for between 8 and 6 points.
A level points? Yes, for reasons best known to UCAS this is not the same as tariff points (so only includes A level performance, not vocational qualifications or grade 8 piano). You get 6 points for an A*, down to 1 point for an E – and only your best three A levels count. So 12 points means three Bs or thereabouts.
The counter story is that this change in behavior hasn’t shifted the overall averages by that much. For high tariff providers the average accepted applicant has 13.9 A level points (down from 14 last year or 14.3 in 2016 – that’s round about AAB. Medium tariff is about BCC (10.4), Lower tariff is near enough CCC (9.4 – up very slightly on the historic average).
Usually I’d suggest that this stasis is down to a regular recalculation of tariff groups – but I know that the last time UCAS allocated providers to groups was back in 2012. We’ve also never been told which providers are in which tariff group – this is a different split to the DfE or OfS variants, unhelpfully. And we don’t get data on A level (or tariff) points by provider, which would offer a much more helpful level of granularity to this point of sector-wide interest.
A peep at provider strategies
There’s been a welcome update to the release of the provider level End of Cycle dataset: previously we used to get offermaking only within a rather vestigial dataset known as “equalities” – 2025 adds the offermaking data plus a range of new equalities parameters to the main provider level release.
For all tariff bands or sector-level data is interesting, the increasing diversity of (and increasing competition within) the sector means that provider-level changes in behavior are by far the most interesting component of this release. The new information means that the chart that you lost your morning to last year is now looking very likely to make you lose your entire day.
This is a complex but powerful dashboard, which shows the difference between the most recent year (2025) of data and a comparator year you can choose (by default last year but you can choose any year since 2019) across two dimensions (you can choose from applications, offers, and accepted applicants for each). I’ve added filters by domicile (UK, international, or all) and subject group (the familiar top level – CAH1 – list).
It’s a lot of data on one chart, so I’ve added a group filter, which by default removes some smaller providers from the display – and there’s a highlighter to help you find a provider of interest.
A dot being further up or further right means that measure has grown between the comparator year and the current year, further down or further left means it has shrunk.
There’s a nearly infinite number of stories to tell from this chart. Here’s some notable ones.
Firstly Canterbury Christ Church University has accepted substantially fewer applicants in 2025 than in 2024. A dig around in the data suggests that decline is focused on UK domiciled applicants studying business subjects, which suggests to me that this shows the end of one or more franchise or partnership arrangements. I asked Canterbury Christ Church University for a comment – nothing yet but I’ll add it if it comes in – I’d imagine that this is the most visible of a wave of providers calculating that the increasing regulatory risk (with both OfS and DfE taking action) is not worth the hassle of running such provision – I’m tentatively pointing at Buckinghamshire New University and Oxford Brookes University as other similar (but smaller) examples).
Not all of the Russell Group is following the same recruitment strategy – there are instances (Nottingham, Glasgow, Cardiff) where fewer applicants have been accepted than in 2024. Some Russell Group providers (for example Leeds, York, Southampton, and Cardiff) have seen fewer applications than in previous years – the first three in that list have nevertheless increased acceptances over last year. Because we can now see the number of offers made using the filters at the top, it is apparent that the entire group (excepting Cardiff and Southampton) made more offers than last year.
League leaders
If you are playing along with the dashboard you’ll have spotted that University College London accepted nearly 2,500 more applicants than last year (after making a genuinely startling 12,000 more offers) . The majority of this increase (2,290 accepted applicants, 10,650 offers) related to international applicants – with growth in pretty much every subject area contributing to this performance.
That’s not the largest growth in accepted applicants, however (it’s the second largest). For the league leaders, we look to the University of Wolverhampton – which accepted an impressive 3,625 extra applicants compared to last year. Unlike UCL, these are all UK-domiciled students, and nearly all (2,970) are studying business subjects. To me, this suggests a new partnership – I asked Wolverhampton about this, and am waiting to hear back.
But who made the most offers in 2025? For international students, it’s UCL and it isn’t even close. But for home students it was the University of Exeter, which made 7,130 more UK domiciled offers this year than last year (a total of 37,515 offers in the 2025 cycle!) across a mix of subject areas. Exeter wasn’t able to get me a comment before publication – I’ll add one if it comes in later.
And I did promise a look at computing recruitment. It is a decline in both applications and acceptances pretty much across the board – with the exception of an 800 student growth in accepted applicants at Bath Spa University. UCL did recruit 40 more students than last year, but this is against a 1,520 decline in applications. There’s still a bit of growth at the University of Manchester, and the University of York – but note also Escape Studios (a growing independent visual effects specialist that was once known as Pearson College, which delivers degrees validated by the Coventry University).
School leavers
I’ve also put together a version of this chart that shows only the recruitment of 18 year olds. The direct path between school or college and university is no longer the dominant one in the UK, and hasn’t been for some time – but in policymaking and political discussions it is still where minds tend to go.
Focusing on UK 18 year olds, we can see that the University of Exeter has grown most spectacularly compared to last year on applications, offers, and acceptances. Large amounts of growth in this part of the market tends to be concentrated in more selective providers, but we can also see credible performances from big civic providers like Nottingham Trent University, Manchester Metropolitan University, and Liverpool John Moores University.
Conversely we can see smaller but notable declines in applications and acceptances from providers including the University of the West of England, Birmingham City University, and the University of East London. The noticeable pattern is that there is no pattern – recruitment among school leavers can go cold anywhere at any time it would seem. And there are some ways around this – both the University of York (up 1,285) and the University of Leeds (up 3,180) upped school-leaver offer making despite a small decline in applications
A sense of the sector
Competition is clearly heating up. For those who have hit on a winning recruitment formula, the challenge becomes a need to ensure that every additional undergraduate gets the high quality experience they have been led to expect. An increase in fee income is almost all going to go to investment in capacity (be that more staff, retaining existing staff, or providing more resources). If your expansion has been into applicant groups you have little experience in teaching, the need to invest rises.
Conversely, for those who have yet to hit upon the way to attract applications reliably there will already have been internal discussions about what needs to be done or what needs to change. Recruitment can and does figure in portfolio review and course revalidation questions: all of which comes down to whether a provider can afford to do what it would like to continue doing. Losing resources or capacity is a very last resort – once you wave goodbye to a course or department it is very difficult to spin back up.
There will also be attention paid to sector trends – the kind of stuff I plotted back in December when we got the first phase of the End of Cycle release. Is it something your provider is doing, or a more general societal change, that means recruitment is growing or shrinking on a particular course. These are difficult, painful conversations, and need careful, considered, responses.